S -Nitrosylation-dependent Inactivation of Akt/Protein Kinase B in Insulin Resistance*

Inducible nitric-oxide synthase (iNOS) has been implicated in many human diseases including insulin resistance. However, how iNOS causes or exacerbates insulin resistance remains largely unknown. Protein S -nitrosy-lation is now recognized as a prototype of a redox-de-pendent, cGMP-independent signaling component that mediates a variety of actions of nitric oxide (NO). Here we describe the mechanism of inactivation of Akt/pro-tein kinase B (PKB) in NO donor-treated cells and diabetic ( db/db ) mice. NO donors induced S -nitrosylation and inactivation of Akt/PKB in vitro and in intact cells. The inhibitory effects of NO donor were independent of phosphatidylinositol 3-kinase and cGMP. In contrast, the concomitant presence of oxidative stress acceler-ated S -nitrosylation and inactivation of Akt/PKB. In vitro denitrosylation with reducing agent reactivated recombinant and cellular Akt/PKB from NO donor-treated cells. Mutated Akt1/PKB (cid:1) (C224S), in which cysteine 224 was substituted by serine, was resistant to NO donor-induced S -nitrosylation and inactivation, indicat-ing that cysteine 224 is a major S -nitrosylation acceptor site. In addition, S -nitrosylation of Akt/PKB

Nitric oxide (NO) 1 is an endogenous cell signaling molecule involved in the regulation of many physiological functions and in the mediation of a variety of pathophysiological processes. NO and NO-related compounds function as both protective and cytotoxic, dependent on the cellular context and the nature of the NO group. The multifaceted actions of the NO group can be classified into two categories: 1) authentic NO-mediated, cGMPdependent, and 2) reactive nitrogen species-mediated, cGMPindependent actions. Nitrosative post-translational modifications, including protein S-nitrosylation and tyrosine nitration, are involved in the cGMP-independent actions. The cGMP-dependent actions play critical roles in a variety of physiological processes, including NO-mediated vasodilation. In contrast, cGMP-independent, nitrosative protein modifications are postulated to be involved in the pathological responses (1)(2)(3)(4).
Nitric-oxide synthases (NOSs) consist of three distinct genes, inducible nitric-oxide synthase (iNOS), endothelial NOS (eNOS), and neuronal NOS (nNOS). NO is generated by iNOS to a much greater extent, to over 1,000-fold, compared with that produced by the constitutive NOSs, eNOS and nNOS (2,5). iNOS and nitrosative stress have been implicated in many human diseases, including insulin resistance (6,7), atherosclerosis (8), inflammation, and neurodegenerative disorders (9). This is largely based on the evidence that iNOS deficiency results in significant amelioration of, or resistance to, these diseases. However, little is known about the molecular mechanisms by which iNOS causes and/or exacerbates these diseases. Furthermore, although previous studies (10,11) showed that iNOS induction or treatment with NO donor inhibited insulin-stimulated glucose uptake, its molecular bases remain unknown.
Protein S-nitrosylation, attachment of nitrosonium ion (NO ϩ ) to cysteine sulfhydryls, has emerged recently as a prototype of cGMP-independent, redox-dependent post-translational modifications (12), which mediate a number of actions of the NO group in various biological processes (13)(14)(15)(16). To date, over 100 proteins have been shown to be S-nitrosylated in vitro, in cultured cells, and in vivo. In many of these proteins, Snitrosylation is associated with functional alterations. This is well exemplified by p21 Ras (17), N-methyl-D-aspartate receptor (18), ryanodine receptor (19), and caspase-3 (20). Recently, protein S-nitrosylation/denitrosylation has been recognized as a regulatory component of signal transduction comparable with phosphorylation/dephosphorylation (21,22). However, in vivo relevance of S-nitrosylation in the pathogenesis of human diseases still remains largely unknown.
Insulin resistance is the major causative factor of type 2 diabetes, a polygenic disease accounting for more than 90% of diabetic patients. Despite intense investigation for a number of years, the molecular mechanisms responsible for insulin resistance remain to be determined. Akt/PKB is a serine/threonine protein kinase, which plays a central role in the metabolic actions of insulin (23), including glucose transport (24) and synthesis of protein and glycogen (25); it is impaired in rodent models of and patients with type 2 diabetes (26 -31), although there is some discrepancy in the literature (32,33). A critical role for Akt/PKB in glucose homeostasis was further corroborated by the evidence that genetic disruption of Akt2/PKB␤ leads to insulin resistance in mice (34). Akt/PKB is activated by phosphatidylinositol 3, 4, 5-triphosphate-dependent phosphorylation at threonine 308 and serine 473 residues. However, little is known about the mechanism that negatively regulates Akt/PKB activity. Therefore, we investigated the role of protein S-nitrosylation in the regulation of Akt/PKB kinase activity and insulin resistance. Here we demonstrate S-nitrosylation-mediated inactivation of Akt/PKB in vitro, in intact cells, and in vivo in genetically obese, diabetic (db/db) mice.

EXPERIMENTAL PROCEDURES
Cell Culture-Mouse C2C12 myoblasts, mouse 3T3-L1 preadipocytes, and monkey COS-7 cells (ATCC, Manassas, VA) were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% heat-inactivated fetal bovine serum (FBS) or bovine calf serum. Human THP-1 myeloid leukemia cells (ATCC) were cultured in RPMI 1640 supplemented with 10% FBS. C2C12 myoblasts were differentiated into myotubes in DMEM supplemented with 2% horse serum for 6 days. Adipocyte differentiation was induced as described previously (35) with minor modifications. Briefly, 3T3-L1 preadipocytes were maintained for 2 days at confluence and then switched to differentiation medium (DMEM supplemented with 10% FBS and supplied with 10 g/ml insulin, 112 g/ml isobutylmethylxanthine, and 1 g/ml dexamethasone). The cells were maintained in differentiation medium for 4 days and then incubated in DMEM supplemented with 10% FBS for an additional 7-10 days. After being serum-starved for 20 h, C2C12 myotubes, 3T3-L1 adipocytes, and COS-7 cells were pretreated with or without S-nitroso-N-acetylpenicillamine (SNAP, 1 mM, Alexis, San Diego, CA) or oxidized SNAP (1 mM) in the presence or absence of carboxy-PTIO (200 M), LY83583 (10 M, Cayman Chemical, Ann Arbor, MI), or glucose oxidase (100 milliunits/ml, Sigma) for 1 h unless otherwise indicated. Then the cells were treated with insulin (100 nM, Sigma) or platelet-derived growth factor (50 ng/ml, Roche Applied Science) or FBS (50%) for 10 min unless otherwise indicated. Oxidized SNAP that no longer generates NO was prepared, as described previously (36), by leaving SNAP, which was dissolved in dimethyl sulfoxide, decomposed at room temperature for 48 h.
Cell Transfection-C2C12 myoblasts and COS-7 cells were transfected using Lipofectamine 2000 reagent (Invitrogen). cDNAs for HAtagged wild-type and constitutively active Akt1/PKB␣ (T308E/S473E) were kindly provided by Dr. D. R. Alessi. cDNAs for mutants of Akt1/ PKB␣ (C60S, C77S, C224S, C296S, C310S, C344S, and C460S), in which cysteine was replaced by serine, were generated by PCR-based, site-directed mutagenesis. The sequences of the primers used to generate mutated Akt1/PKB␣ are provided in supplemental Table I. At 48 h after the transfection with constitutively active Akt1/PKB␣ (T308E/ S473E), C2C12 myoblasts were treated with SNAP (1 mM) for 1 h or S-nitrosoglutathione (GSNO, 1 mM, Cayman Chemical) for 3 h in the presence of 10% FBS. COS-7 cells were transfected with Myc-or FLAGtagged wild-type and cysteine-to-serine mutants of Akt1/PKB␣. At 6 h after the transfection, the medium was replaced with serum-free medium. After COS-7 cells were serum-deprived for 20 h, and the cells were treated with or without SNAP (1 mM) for 1 h in the presence of carmustine (80 M, Sigma).
In COS-7 cells transfected with wild-type Akt1/PKB␣ or mutated Akt1/PKB␣ (C224S), following serum starvation for 20 h, the cells were treated with 50% FBS for 10 min. Immunoprecipitated wild-type and mutated Akt1/PKB␣ were incubated with or without SNAP (1 mM) at 4°C for 1 h before in vitro Akt/PKB kinase assay was performed.
In vitro kinase reaction was performed in kinase buffer (50 mM HEPES, pH 7.4, 10 mM MgCl 2 , 1 mM CaCl 2 , 1 mM dithiothreitol, and 0.1 mM sodium vanadate) containing 1 Ci of [␥-32 P]ATP (PerkinElmer Life Sciences) at 30°C for 10 min. Incorporation of ␥-32 -phosphate into Crosstide was measured with a scintillation counter. In the indicated experiments, the immunoprecipitates were incubated with or without dithiothreitol (20 mM) for 20 min at 30°C. During all experimental procedures, exposure of the samples to light was minimized.
PI 3-Kinase Assay-Insulin-stimulated PI 3-kinase activity was evaluated in C2C12 myotubes. Following serum starvation for 20 h, the cells were pretreated with or without SNAP (1 mM) for 1 h and then stimulated with insulin (100 nM) for 10 min. PI 3-kinase activity in immunoprecipitates with anti-phosphotyrosine (Santa Cruz Biotechnology) was measured, as described previously (38). Briefly, in vitro phosphorylation reaction of phosphatidylinositol (Sigma) was performed in PI 3-kinase buffer (100 mM Tris-HCl, pH 7.5, 100 mM NaCl, 1 mM EDTA, 15 mM MgCl 2 , 70 M ATP) containing 10 Ci of [␥-32 P]ATP at 37°C for 10 min. The reaction was stopped by the addition of 6 N HCl. The phospholipids were extracted with chloroform/methanol (1:1, v/v). The organic phase, containing the labeled PI 3-kinase products, was separated by silica gel TLC plates (Sigma). The plate was then developed in CHCl 3 /CH 3 OH/H 2 O/NH 4 OH (60:47:11:3.2), dried, and visualized by autoradiography.
S-Nitrosylation of recombinant Akt1/PKB␣ was evaluated by UVvisible spectrophotometric analysis, as described previously (40). Recombinant active Akt1/PKB␣ protein (Upstate Biotechnology, Inc.) was incubated with or without SNAP (1 mM) for 1 h at room temperature. After passing Sephadex G-50 columns, recombinant Akt1/PKB␣ was incubated with or without dithiothreitol (20 mM) for 20 min at 30°C. S-Nitrosylation was detected as a peak at ϳ340 nm.
Animals-Male, genetically obese, diabetic (db/db) mice and lean wild-type (C57BL/6J) mice were purchased from The Jackson Laboratory (Bar Harbor, ME). Diabetic (db/db) mice were backcrossed onto C57BL/6J five generations. The Institutional Animal Care Committee approved the study protocol. The animal care facility is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care. At 10 weeks of age, following an overnight fast, the animals were anesthetized with an intraperitoneal injection of pentobarbital sodium (60 mg/kg body weight). Skeletal muscle (gastrocnemius) was taken at 8 min after insulin (5 units/kg BW, Humulin R, Lilly) or saline was injected via the portal vein. The muscle tissues were homogenized in buffer D (50 mM HEPES, pH 7.5, 150 mM NaCl, 2 mM EDTA, 1% Nonidet P-40, 10% glycerol, 10 mM sodium fluoride, 2 mM sodium vanadate, 1 mM PMSF, 10 mM sodium pyrophosphate, 5 g/ml aprotinin, 5 g/ml leupeptin, 0.5 g/ml pepstatin) using Polytron PT-MR 3000 (KINEMATIKA AG, Littau, Switzerland). Blood glucose was measured using an Elite glucometer (Bayer, Elkhart, IN).
Statistical Analysis-The data were compared using one-way analysis of variance followed by Scheffe's F-test. The null hypothesis was rejected when p Ͻ 0.05. Values are reported as means Ϯ S.E.

NO Donor Inhibits Akt/PKB Activation in Intact Cells-To
investigate the mechanisms underlying iNOS-involved insulin resistance, we first examined the effects of NO donor on insulin-stimulated activation of Akt/PKB in cultured skeletal muscle cells and adipocytes. Pretreatment with the NO donor SNAP (1 mM) for 1 h resulted in inhibition of insulin-stimulated Akt/PKB activation in mouse C2C12 myotubes (Fig. 1, A-C and E) and mouse 3T3-L1 adipocytes (data not shown). SNAP did not affect the protein expression of Akt/PKB (Fig. 1, B-D). Oxidized SNAP, which no longer generates NO, failed to inhibit Akt/PKB activation (Fig. 1B). NO scavenger carboxy-PTIO prevented the inhibitory effects of NO donor. On the other hand, the inhibitor for soluble guanylate cyclase, LY83583, failed to block the effects of SNAP (Fig. 1B). This finding indicates that the inhibitory effect of NO donor is cGMP-independent. The inhibitory effect of NO donor is not limited to insulin-stimulated Akt/PKB activation. SNAP inhibited FBS-and plateletderived growth factor (PDGF)-induced activation of Akt/PKB in C2C12 myotubes (Fig. 1C). The basal Akt/PKB activity in human myeloid leukemia THP-1 cells cultured in the medium supplemented with 10% FBS was also inhibited by SNAP (Fig.  1D). In contrast, SNAP did not inhibit insulin-induced ERK activation, which is a signaling cascade parallel to the Akt/PKB pathway downstream of the insulin receptor (Fig. 1E).
Insulin activates PI 3-kinase-Akt/PKB via insulin receptor substrate (IRS)-1 and -2 in skeletal muscle and adipocytes. On the other hand, IRSs are not required for PDGF-induced activation of PI 3-kinase-Akt/PKB. THP-1 cells are deficient in IRS-1 and IRS-2 (41). The findings that NO donor effectively reduced PDGF-induced Akt/PKB activation in C2C12 myotubes (Fig. 1C) and Akt/PKB activity in THP-1 cells (Fig. 1D) suggest that the NO donor can exert its inhibitory effects on Akt/PKB in a manner independent of IRS-1 and IRS-2.
Inactivation of Akt/PKB by NO Donor Is PI 3-Kinase-independent-Consistent with the inhibition of Akt/PKB activation, SNAP blocked insulin-stimulated phosphorylation of endogenous substrates of Akt/PKB, FOXO1, and Bad ( Fig. 2A). In contrast, insulin-stimulated phosphorylation of threonine 308 and serine 473 residues of Akt/PKB, which is PI 3-kinase-dependent and leads to the activation of Akt/PKB, was not af-fected by SNAP. In accordance with this finding, insulin-stimulated PI 3-kinase activation was not impaired by the treatment with SNAP for 1 h either (Fig. 2B). These observations indicate that the inhibitory effect of NO donor is PI 3-kinase-independent. This is further confirmed by the evidence that NO donors SNAP and GSNO inhibited the constitutively active mutant of Akt/PKB (Fig. 2C), in which threonine 308 and serine 473 residues were replaced by glutamic acid. Taken together, these findings indicate that the inactivation of Akt/PKB by SNAP is phosphorylation status-independent.
S-Nitrosylation/Denitrosylation Is Associated with Inactivation/Reactivation of Akt/PKB by NO Donor-PI 3-kinase-independent inactivation of Akt/PKB is suggestive of a direct inhibition of the kinase by the NO group. Therefore, to address whether the NO donor can directly inactivate Akt/PKB, we examined the effects of SNAP on recombinant active Akt/PKB protein in vitro. SNAP and GSNO, but not oxidized SNAP, directly inactivated recombinant Akt/PKB in vitro (Fig. 3A, and data not shown). Most importantly, the in vitro incubation with . SNAP inhibited FBS (50%)-and PDGF-stimulated (50 ng/ml) Akt/PKB activation (C), as well as insulin-stimulated activity in C2C12 myotubes. The basal activity of Akt/PKB was also reduced when THP-1 cells were exposed to SNAP (1 mM) for the indicated times (D). In contrast, SNAP did not decrease insulin-stimulated ERK activation in C2C12 myotubes (E). On the other hand, SNAP did not affect the protein expression of Akt/PKB (lower panels of B-D). Akt/PKB and ERK were immunoprecipitated (IP) to assess kinase activities by in vitro phosphorylation assay using Crosstide (A-D) or histone H2B (H2B) and myelin basic protein (MBP) (E) as substrates, respectively. dithiothreitol (20 mM) restored the kinase activity of the recombinant protein (Fig. 3A). Immunoprecipitated Akt/PKB from NO donor-treated C2C12 myotubes was also reactivated by dithiothreitol (Fig. 3C). Inhibition of the kinase activity by NO donor and its reversal by dithiothreitol correlates quite well with simultaneous S-nitrosylation and denitrosylation of the kinase (Fig. 3, B and C, lower panel).
Hydrogen peroxide is capable of activating Akt/PKB in the absence of insulin or other growth factors (42). We found that in vitro incubation with SNAP (1 mM) for 1 h inactivated hydrogen peroxide (1 mM)-induced activity of the kinase by 70% (data not shown). Taken together, these findings clearly indicate that the NO donor directly inactivates Akt/PKB by Snitrosylation in vitro and in intact cells.
Oxidative Stress Facilitates NO Donor-induced Akt/PKB Inactivation-In the pathological conditions associated with iNOS induction, which include type 2 diabetes, there co-exists increased oxidative stress. Moreover, oxidative stress is known to accelerate protein S-nitrosylation (43,44). Therefore, we investigated the effects of concomitant oxidative stress in NO donor-induced inhibition of Akt/PKB activation. In the absence of concomitant oxidative stress, the inhibitory effects of SNAP on insulin-stimulated Akt/PKB activation were observed at concentrations of 33 M and higher in C2C12 myotubes (Fig.  4A). On the other hand, in the presence of glucose oxidase (100

FIG. 2. The inhibition of Akt/PKB by NO donor is PI 3-kinaseindependent.
A, immunoblot analysis revealed that SNAP inhibited insulin-stimulated phosphorylation of FOXO1 and Bad, which are endogenous substrates of Akt/PKB, in C2C12 myotubes. In contrast, insulin-stimulated phosphorylation of Akt/PKB at threonine 308 and serine 473 was not affected. B, insulin-stimulated PI 3-kinase activity was not decreased by SNAP in C2C12 myotubes. C, SNAP and GSNO reduced the activity of ectopically expressed, constitutively active mutant of Akt/PKB (HA-tagged Akt1/PKB␣ T308E/S473E) in C2C12 myoblasts.

FIG. 3. NO donor-induced S-nitrosylation and inactivation of Akt/PKB and its reversal by reducing agent. Recombinant active Akt1/PKB␣ protein was inactivated by SNAP (1 mM) but not by oxidized SNAP (1 mM). In vitro incubation with dithiothreitol (DTT) (20 mM) restored the activity of SNAP-treated Akt1/PKB␣ (A). SNAP (1 mM) induced S-nitrosylation of recombinant Akt1/PKB␣ (B) and endogenous
Akt/PKB in C2C12 myotubes (C). S-Nitrosylation was detected by the peak at ϳ340 nm using an UV-visible spectrophotometer (B), and by the biotin-switch method as described under "Experimental Procedures" (C). In vitro incubation with dithiothreitol (20 mM) reversed SNAP (1 mM)-induced S-nitrosylation of recombinant Akt1/PKB␣ (B) and immunoprecipitated Akt/PKB from SNAP-treated (1 mM) C2C12 myotubes (C, lower panel). The reversal of S-nitrosylation was associated with the restoration of insulin-stimulated Akt/PKB activity in SNAP-treated C2C12 myotubes (C). milliunits/ml), which generates hydrogen peroxide, SNAP effectively inhibited insulin-stimulated Akt/PKB activation at concentrations more than 1 order of magnitude lower than those in the absence of glucose oxidase (Fig. 4B). In agreement with previous observations of enhanced S-nitrosylation of other proteins by oxidative stress (43,44), glucose oxidase facilitated S-nitrosylation of Akt/PKB by NO donor (Fig. 4C). Similarly, SNAP-induced S-nitrosylation of Akt/PKB was facilitated by the co-administration of carmustine (80 M) (data not shown), a specific inhibitor for glutathione reductase that is an important anti-oxidant molecule used to maintain a reducing condition in cells. Most importantly, the accentuation of NO donorinduced Akt/PKB inactivation by glucose oxidase paralleled the facilitated S-nitrosylation of the kinase by concomitant oxidative stress. This provides further support for the involvement of S-nitrosylation.
Cysteine 224 Is a Major S-Nitrosylation Acceptor Site-Human Akt1/PKB␣ contains seven cysteine residues. To determine a cysteine residue sensitive to NO donor-induced S-nitrosylation and inactivation, we generated mutants of human Akt1/PKB␣, in which each of the seven cysteine residues was replaced by serine. Treatment with SNAP (1 mM) in the presence of carmustine (80 M) induced a marked S-nitrosylation of wild-type Akt1/PKB␣ (Fig. 5A). By contrast, Akt1/PKB␣ (C224S), in which cysteine 224 was substituted by serine, was resistant to S-nitrosylation upon exposure to the NO donor SNAP (Fig. 5A). However, other mutants of Akt1/PKB␣ (C60S, C77S, C296S, C310S, C344S, and C460S) were S-nitrosylated by the NO donor to the extent comparable with that observed in wild-type Akt1/PKB␣ (Fig. 5A). Furthermore, SNAP failed to decrease the activity of mutated Akt1/PKB␣ (C224S), whereas wild-type Akt1/PKB␣ was significantly inactivated (Fig. 5B).

Increased S-Nitrosylation of Akt/PKB in Skeletal Muscle of Diabetic (db/db) Mice-To address in vivo biological relevance of S-nitrosylation-involved
Akt/PKB inactivation, we examined insulin-stimulated Akt/PKB activation in skeletal muscle of genetically obese, diabetic (db/db) mice. Fasting blood glucose levels of wild-type and db/db mice were 146.6 Ϯ 8.0 and 278.6 Ϯ 22.3 mg/dl, respectively. In agreement with the previous finding (28), insulin-stimulated Akt/PKB activity in db/db mice was decreased to 52% that in wild-type mice, as judged by the immune complex kinase assay (Fig. 6A). Although there was no difference in Akt/PKB expression, insulin-stimulated phosphorylation of Akt/PKB at threonine 308 in db/db mice was decreased to 67% those in wild-type mice (Fig. 6B). Thus, the extent of the reduction in insulin-stimulated Akt/PKB activity in db/db mice (Fig. 6A) appeared to be greater relative to the decrease in insulin-stimulated phosphorylation of Akt/PKB in db/db mice (Fig. 6B). In fact, the ratio of insulin-stimulated activity of immunoprecipitated Akt/PKB (Fig. 6A) to insulinstimulated Akt/PKB phosphorylation (Fig. 6B) was significantly lower in db/db mice than in wild-type mice (Fig. 6C). S-Nitrosylation of Akt1/PKB␣ was detected by the biotin-switch method as described under "Experimental Procedures." SNAP treatment induced S-nitrosylation of wild-type Akt1/PKB␣. However, SNAP-induced S-nitrosylation was significantly attenuated in Akt1/PKB␣ (C224S) (A). In contrast, other mutants of Akt1/PKB␣ were substantially S-nitrosylated by SNAP (A). B, the kinase activity of wild-type Akt1/PKB␣ and mutated Akt1/PKB␣ (C244S) was assessed by in vitro phosphorylation assay. The kinase activity of wild-type Akt1/PKB␣ was reduced by SNAP. However, SNAP failed to affect the kinase activity of mutated Akt1/PKB␣ (C244S).
Most importantly, the in vitro incubation with dithiothreitol (20 mM) restored partially, but significantly, the activity of immunoprecipitated Akt/PKB from skeletal muscle of db/db mice to 73% that in wild-type mice (Fig. 6A). By contrast, the incubation with dithiothreitol did not significantly increase Akt/PKB activity in wild-type mice. Of note, after the incubation with dithiothreitol, no significant difference was found in the ratio of insulin-stimulated activity of immunoprecipitated Akt/PKB to insulin-stimulated Akt/PKB phosphorylation between db/db mice and wild-type mice (Fig. 6C). Consistent with the reversal of Akt/PKB activity by dithiothreitol, S-nitrosylated Akt/PKB was significantly increased in skeletal muscle of db/db mice compared with wild-type mice (Fig. 6D). DISCUSSION We found that NO donors inactivated Akt/PKB, a pivotal kinase in the metabolic actions of insulin, both in vitro and in intact cells with simultaneous S-nitrosylation of Akt/PKB. The direct inactivation of Akt/PKB was S-nitrosylation-dependent and PI 3-kinase-independent. Moreover, S-nitrosylated Akt/ PKB was increased in genetically obese, diabetic (db/db) mice. Most importantly, the reversal of S-nitrosylation of Akt/PKB by reducing agents in vitro reactivated NO donor-treated recombinant Akt/PKB and endogenous Akt/PKB in NO donortreated cells. These results suggest that the activity of Akt/ PKB is regulated by S-nitrosylation/denitrosylation and therefore reversal of S-nitrosylation of Akt/PKB may serve as a potential therapeutic target for reversal of attenuated metabolic actions of insulin.
Although previous studies implicated iNOS and oxidative stress in insulin resistance (5,6,9,10), the molecular mechanisms underlying iNOS-and oxidative stress-involved insulin resistance remains largely unknown. The present study suggests that iNOS and oxidative stress may contribute to the development and/or exacerbation of insulin resistance, at least in part, by S-nitrosylation-mediated inactivation of Akt/PKB. Substitution of cysteine 224 of human Akt1/PKB␣ by serine conferred resistance to NO donor-induced S-nitrosylation and inactivation of Akt/PKB (Fig. 5). This finding indicates that a major S-nitrosylation acceptor site in Akt1/PKB␣ is cysteine 224, which is conserved from Caenorhabditis elegans to humans, and also among Akt1/PKB␣, Akt2/PKB␤, and Akt3/PKB␥.
Of interest, in contrast to mitigation of the diseases by iNOS deficiency, the deficiency of constitutive NOSs (eNOS and nNOS) is associated with induction and/or deterioration of the disease states in most cases, if not all. This is the case with insulin resistance. Although iNOS deficiency protects from high fat diet-induced insulin resistance (6), targeted gene disruption of eNOS or nNOS is associated with insulin resistance (45)(46)(47). Likewise, although iNOS-specific inhibitor prevented lipopolysaccharide-induced insulin resistance (7), nonselective NOS inhibitors were shown to cause insulin resistance (48), even though the controversial results were also reported (49,50) regarding the effects of nonselective NOS inhibitors. The molecular basis accounting for this apparent discrepancy, however, remains an open question.
The Janus-faced nature of the NO group is assumed to be largely dependent on redox chemistry, namely nitrosative protein modification-and cGMP-dependent actions that are mediated by reactive nitrogen species and authentic NO, respectively. Oxidative stress promotes the generation of reactive nitrogen species, such as nitrosonium ion (NO ϩ ) equivalent and peroxynitrite (OONO Ϫ ), and facilitates nitrosative post-FIG. 6. Involvement of S-nitrosylation in impaired Akt/PKB activation in skeletal muscle of genetically obese, diabetic (db/db) mice. A, insulin-stimulated Akt/PKB activation was impaired in skeletal muscle of genetically obese, diabetic (db/db) mice compared with wild-type (WT) mice, whereas there was no difference in the protein expression of Akt/PKB. In vitro incubation with dithiothreitol (DTT) (20 mM) up-regulated significantly the reduced insulin-stimulated Akt/PKB activity in diabetic (db/db) mice but not in wild-type mice. B, insulinstimulated phosphorylation was attenuated in skeletal muscle of diabetic (db/db) mice compared with wild-type mice. C, the ratio of insulinstimulated Akt/PKB activity to the level of phosphorylated Akt/PKB was lower in diabetic (db/db) mice than in wild-type mice, when immunoprecipitated Akt/PKB was not incubated with dithiothreitol (20 mM). In vitro incubation with dithiothreitol (20 mM) reversed the decrease in diabetic (db/db) mice in the ratio of insulin-stimulated Akt/PKB activity to phosphorylated Akt/PKB. D, S-nitrosylation of Akt/PKB in skeletal muscle was assessed by the biotin-switch method as described under "Experimental Procedures." S-Nitrosylated Akt/PKB was significantly increased in diabetic (db/db) mice compared with wild-type mice. translational modifications, including S-nitrosylation and tyrosine nitration. In many cases, iNOS is associated with nitrosative post-translational modifications, because iNOS induction is accompanied by oxidative stress, and much higher concentrations of the NO group are needed for nitrosative protein modifications compared with NO concentrations required to increase cGMP production. Although low nanomolar concentrations of NO donors are sufficient to elicit cGMP-dependent signals, 50 -1000 M of NO donors is required for S-nitrosylation-mediated alterations of protein function in cultured cells (51)(52)(53). A question has been raised about the pathophysiological relevance of S-nitrosylation observed in cells treated with pharmacological doses (50 -1000 M) of NO donors. Recently, however, facilitation of S-nitrosylation under hypoxic conditions (54) or by concomitant oxidative stress (43,44) has been proposed to account for S-nitrosylation in tissues in vivo, where GSNO concentration can be up to a low micromolar concentration. Consistent with the previous findings of the acceleration of S-nitrosylation by oxidative stress (43,44), we found that oxidant, glucose oxidase, enhanced the inhibitory effects of NO donor on Akt/PKB activity and also S-nitrosylation of the kinase (Fig. 4). One can reasonably speculate that increased S-nitrosylation of Akt/PKB in db/db mice (Fig. 6) may be attributable, at least in part, to elevated oxidative stress in diabetic mice.
In vitro incubation with dithiothreitol up-regulated significantly the activity of Akt/PKB immunoprecipitated from skeletal muscle of db/db mice (Fig. 6A). However, in vitro incubation with dithiothreitol did not completely revert the reduced insulin-stimulated Akt/PKB activity in db/db mice to the level in wild-type mice. This indicates that there exists other molecular mechanism(s) underlying attenuated PI 3-kinase-Akt/ PKB activation in db/db mice, in addition to S-nitrosylation of Akt/PKB. Insulin-stimulated phosphorylation of Akt/PKB, which mediates PI 3-kinase-dependent activation of Akt/PKB, was also decreased in db/db mice compared with wild-type mice (Fig. 6B). Notably, the ratio of insulin-stimulated activity of immunoprecipitated Akt/PKB to insulin-stimulated phosphorylation of Akt/PKB was lower in db/db mice than in wildtype mice, but this decrease was completely reversed by in vitro incubation with dithiothreitol (Fig. 6C). These findings suggest that both S-nitrosylation of Akt/PKB and the defect(s) at the level and/or upstream of PI 3-kinase, the latter of which is reflected by impaired phosphorylation of Akt/PKB, contribute in concert to the attenuated insulin-stimulated Akt/PKB activity in skeletal muscle of db/db mice.
Likewise, the NO donor seems to impair insulin-stimulated activation of PI 3-kinase-Akt/PKB at multiple levels of insulin signaling cascade. Our preliminary results demonstrated that the NO donor GSNO (1 mM) can also reduce IRS-1 expression in a time-dependent fashion at times well beyond that used in this study (1 h). 2 It is important to note, however, that the inhibitory effects of the NO donor on Akt/PKB is independent of the effects on IRS-1 or IRS-2; the treatment with the NO donor for 1 h did not cause a decrease in the expression of IRS-1 or IRS-2 (supplemental Fig. 1) but did reduce Akt/PKB activity. Notably, SNAP did also inactivate PDGF-stimulated Akt/PKB activation (Fig. 1C) that is independent of IRSs. Moreover, SNAP reduced Akt/PKB activity in THP-1 cells that are deficient in IRS-1 and IRS-2 (41).
The present study demonstrated the selective inhibition of Akt/PKB by NO donor with no inhibitory effect on ERK activation (Fig. 1E). The PI 3-kinase-Akt/PKB and ERK pathways are two major signal transduction cascades activated by insulin; each plays an important role in the metabolic and mitogenic actions of insulin, respectively. Our results are consonant with the previous findings that the PI 3-kinase-Akt/PKB pathway was selectively impaired with intact ERK activation in rodent models of insulin resistance and type 2 diabetes (28,(55)(56)(57)(58). This selective impairment of the PI 3-kinase-Akt/PKB pathway combined with hyperinsulinemia has been postulated to underlie the pathogenesis of diabetic complications such as macroangiopathy. Gene disruptions of iNOS and iNOS inhibitors are known to protect from the development of diabetic complications including vasculopathy (59,60). Therefore, the present study suggests the possibility that S-nitrosylation may contribute to diabetic complications via selective impairment of insulin signaling. This possibility deserves further investigation.